The hypothesis to be tested is that the use of an antioxidant cocktail (vitamin E, C and B-carotene) will reduce indirect measures of lipoprotein oxidation in subjects with diabetes who are known to have increased oxidant stress.